It's amazing what you can see when you refuse to open your eyes -- or need to talk your book. Take, for example, Marc Andreessen's article in the WSJ titled "Why Software is Eating the World." I became skeptical when this line appeared in the introduction:
And, perhaps most telling, you can't have a bubble when people are constantly screaming "Bubble!"
It would seem this is what passes for conventional wisdom in Silicon Valley: utter denial of facts on that grounds that if people recognize them, they can't be true. It's cute, like they're really trying to understand finance but haven't quite got it right. Maybe someone told them that due to efficient markets, if everyone knows something it will be completely priced in. Like anyone with a vested interest, these people have decided that their markets qualify as efficient and so the axiom must hold (or, to the extent that markets are inefficient, it's in such a way that they will profit enormously). Unfortunately, the entire statement represents terrible investment logic, a perversion of the fact that we can not know the magnitude of a bubble except in retrospect. That says nothing about the existance of a bubble. It's critical to remember that we can -- and good investors, by definition, do -- know if there's a bubble before it bursts.
But Marc's statement itself isn't even true. Hardly anyone is screaming "Bubble!" In fact, methinks the technology investors doth protest too much. The most telling sign of all is that the bubble-apologists cry much louder than their accusers! For every journalist who dares write that there is a bubble, two write articles saying that the low current prices of 2011 IPO's prove the opposite.
But onto the real absurdity of the article. Specifically, the very next sentence:
But too much of the debate is still around financial valuation, as opposed to the underlying intrinsic value of the best of Silicon Valley's new companies.
I choked when I read that, having written a post a few weeks ago arguing exactly the opposite. There is no debate about financial valuation of software companies -- they have it. They have way too much of it. On the other hand, what justifies it? Most value investors -- who define holding periods in years, rather than seconds -- regard this industry with skepticism.
Don't worry, though, because Marc provides examples to support his idea.
The first "software company" he mentions is Amazon. This is a confusing choice because it has a P/E of close to 80, and would seem to run counter to his argument that there are "all-time low price/earnings ratios for major public technology companies." Also, it's not a software company. Software services are a growing fraction of their revenue, but Amazon ships physical products to physical people. They have a massive software architecture to support that, and major software projects like the Kindle, but this is not a software company. The recent news that ebook sales are surpassing those of physical books means they are closer to this definition, but still far from it. Not a software company.
Next up: Netflix. Another "software company" that ships physical products to physical people. Yes, they recently rolled out an on-demand service, but its selection remains limited when compared to the physical catalogue. Not a software company.
Next up: music companies. First, Apple's iTunes -- a software product completely devoid of value when not linked to the (only) mobile devices it supports. For all Marc's talk about how Apple is the most valuable company in the world, I wonder if he's thought about how much of that value comes from software. The answer is: very little. His other two examples, Spotify and Pandora, have both failed thus far to show a profit. That hasn't stopped the latter from IPO'ing, however.
Next up: video game companies Rovio and Zynga. No arguments here -- these companies successfully use software platforms to deliver content that people want. Neither of them pushes the state of the art or has any remotely defensible position because their products are commoditzed, but software companies they are.
Next up: Pixar. Pixar makes software, sure, but how can they be a "software company"? Their revenue is solely derived from me getting in my car and walking into a movie theatre. Is Toy Story a software product (not a product of software)? If it is, I have to take back what I said about video game companies, because Marc says that "traditional" game companies like EA are failing -- but their products are as much software as Pixar's if not more, since you don't have to leave home to use them!
Next up: Photography. Marc is really starting to reach here -- he's comparing revenue-lacking software sharing companies to a hundred year old photographic film and paper company (Kodak).
Next up: Local companies. Sorry, now he's stretching. Comparing Google to companies like Groupon and Foursquare, whose accounting practices of "earnings before costs" make them the laughingstock of the investment industry.
We're not done: Skype is next, and Marc uses the $8.5 billion price tag to justify his choice (despite earlier claiming that valuations are at an all time low). No mention that the next best bid was half that. Skype is followed by LinkedIn, another 2011 IPO that has a long way to go before demonstrating real success.
You get the idea. The balance of the article argues that because other physical industries like Wal-Mart and oil/gas use software, they represent the triumph of software over more traditional methods.
No one is arguing that software doesn't make things better, more efficient, cheaper. No one is arguing that software isn't growing tremendously as an integral part of modern corporations. But the bubble apologists continue to ignore the actual state of software. Merely mentioning Groupon, LinkedIn, Pandora, etc. doesn't make them successful -- and yet article after article continues to be written about them completely devoid of justifying arguments. Cloud companies like Salesforce trade at ratios of over five hundred times earnings!
The companies Marc describes -- at least, the successful ones -- could be best described as software-aided companies, or software-centric companies. Most of his choices exhibit strong selection bias, or rose to prominence in hardware before moving to software. Google -- which he mentions as a competitor of the software companies! -- is one of the only good counterexamples.
So here's my question: where does the hardware to support all this software come from? I'd like to see that company, because they must be doing unbelievably well. Who knows, maybe the next great "software company" will be an infrastructure firm in disguise. After all, if they have a website, they must be in software!